Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

Summary of the workshop


Scientific experts in fields concerned with the health and well-being of children and their mothers recognize breastfeeding as the best way to nourish infants and to promote the post-partum and possibly the longer-term health of women. This consensus led the world community more than 20 years ago to recommend that infants be breastfed exclusively for four to six months, with continued breastfeeding for two years or longer.


Benefits to the infant

The benefits to the infant fall into three broad categories: nutritional, immunologic, and behavioural. Reviews of the world's scientific literature by the Working Group reaffirmed the strength of the conclusions regarding the nutritional and immunologic benefits and led to an acknowledgement of the potential for significant behavioural advantages.

Nutritional benefits

It is clear that the nutritional benefits to infants extend through the periods of exclusive and partial breastfeeding and possibly beyond the latter period. The high nutritional value of human milk conferred by the high bioavailability of its nutrients, the balance of specific nutrients, and other characteristics is of significant advantage to all infants. It is of most value, however, to infants of families with low economic resources. Human milk substitutes accessible to those infants generally are unsafe because of their inferior nutritional quality and frequent contamination with potentially fatal infectious agents.

The bacteriologic safety and high nutritional value of human milk argue strongly for maximizing the intake of human milk, especially among infants of the poor. This is especially important when other foods are added to the infant's diet. Added foods are of inferior quality, are often bacteriologically contaminated, and generally displace human milk, thereby raising the risk of infection and malnutrition in children living in unsafe environments. Although there are recent data that support the extension of exclusive breastfeeding beyond six months among some infants, the published data remain too limited to conclude that the period of exclusive breastfeeding should be extended universally beyond the current recommended period of four to six months. There is no controversy, however, regarding the important role of human milk in supplying essential nutrients to the child during the period of mixed feeding.

The major possible exceptions to the full adequacy of exclusive breastfeeding relate to vitamins K, D, and B12. Although there is some disagreement, most medical scientists continue to recommend vitamin K supplementation at birth for all infants, regardless of feeding mode, to prevent haemorrhagic problems in the newborn. Infants with limited exposure to sunlight, or those whose mothers had low vitamin D stores because of low vitamin D intake or limited sun exposure, should receive vitamin D supplements. Infants of strict vegetarians who do not eat eggs or milk products also run the risk of vitamin B12 deficiency and should be provided with an exogenous source of this essential nutrient.

Maternal malnutrition also may result in abnormally low levels of some nutrients in human milk. Generally, however, even women living under very harsh conditions will provide sufficient milk of adequate quality to breastfeed infants exclusively for four to six months. More importantly, circumstances that lead to maternal malnutrition almost uniformly result in malnutrition and serious infectious morbidity among non-breastfed infants. The mortality rates of non-breastfed infants in these circumstances are estimated to be 12 times higher than those of breastfed infants.

The contamination of human milk by xenobiotics may present safety concerns in some circumstances. Situations in which environmental pollutants, such as heavy metals and organohalides, may contaminate human milk should be evaluated carefully; however, attention must always be given to the benefits and risks presented by the exclusion of human milk and the use of its substitutes. In these circumstances exposure of the infant obviously begins in utero, thus making it more imperative to attend to the contamination of the environment.

The major biological xenobiotic of concern is the human immunodeficiency virus (HIV), which is responsible for the acquired immunodeficiency syndrome (AIDS). Although it is clear that human milk may carry HIV, controversy remains as to the conditions that determine the infectivity of HIV in human milk. There is no controversy, however, that the benefits of human milk are much greater than the risks presented by HIV in areas characterized by high rates of infant mortality and malnutrition.

Immunologic benefits

In addition to serving as the most reliable, safe, and nutritious food for infants, human milk provides unique immunologic benefits, which result in decreased rates of infection and other desirable outcomes. In the recent past, scientific evidence was limited to the likelihood that human milk conferred passive protection; that is, protection against infectious disease occurred because of the direct interaction between specific milk components and potential pathogens that threaten the infant. However, there was little evidence to support the hypothesis that human milk alters the development of the infant's immune system to provide active as well as passive protection. Data reviewed by the Working Group support the idea that both active and passive mechanisms likely account for the decreased infectious morbidity observed in infants in both economically developing and fully industrialized nations. Of particular interest were data from the United Kingdom demonstrating that breastfeeding for at least 13 weeks had protective effects against gastrointestinal and, to a lesser degree, respiratory infections that lasted beyond weaning.

Recent studies have demonstrated that breastfeeding enhances responses by the infant to infectious challenges. Enhanced responses were noted following the administration of parenteral vaccination with diphtheria and tetanus toxoids and Hib (Haemophilus influenzae type b)-protein conjugate, oral poliovirus, and "natural" infections with respiratory syncytial virus. The mechanisms responsible for these responses are the subject of intensive investigations. It is likely that both the high nutritional quality of human milk and its complex immune components (e.g., various growth factors, cytokines, and antiinflammatory factors) are responsible for the improved immune function of breastfed infants.

The combined nutritional and immunologic protective effects of human milk against diarrhoeal disease result in a reduced incidence and severity of the disease. As a consequence, breastfeeding protects strongly against diarrhoeal mortality, especially in young infants. Although partial breastfeeding is protective, maximal protection is achieved with exclusive breastfeeding. A reduction in the incidence of respiratory disease is less clearly established, but breastfeeding does appear to reduce the severity of respiratory illness, as reflected by hospitalization rates and mortality. There also is fairly consistent evidence that breastfeeding protects against otitis media, but the effect is less than that seen for diarrhoeal diseases. Evidence for protection against other infectious diseases is less clear, but nonetheless suggestive. Theoretical mathematical projections based on data obtained from the World Health Organization indicate that a 40% reduction in the prevalence of non-breastfeeding would result in a 50% reduction in respiratory deaths and a 66% reduction in diarrhoeal deaths worldwide in children 18 months old or younger.

Evidence also was reviewed which suggests that the immunologic benefits may last for a longer term. Investigators have reported that for years after breastfeeding has ceased, breastfed infants have a significantly lower risk than bottle-fed infants of developing type I diabetes, Crohn's disease, and Iymphomas in childhood.

Behavioural benefits

Behavioural benefits are more difficult to document. Although it is highly plausible that specific constituents of human milk enhance the infant's neural development and that suckling at the breast promotes desirable emotional ties between mother and infant, objective experimental evidence in support of the hypothesis that breastfeeding directly enhances the infant's behavioural development is limited. The usefulness of most published investigations is restricted by inadequate study designs, inappropriate evaluation tools available to or selected by the researchers, and an overly narrow focus on developmental outcomes, such as IQ scores and psychomotor indices. The narrowness of the focus excludes consideration of interactions between feeding mode and other potentially important modulators of behavioural development (such as reductions in morbidity) and disregards the processes that underlie development. Furthermore, very little attention has been given to the alternative possibility that breastfeeding may limit mental development through, for example, the transfer to the child of toxic substances in milk. This alternative is complicated by the confounding likelihood that the infant's exposure to toxicants is initiated during gestation, the period of maximal vulnerability.

Even when these caveats are acknowledged, previously breastfed children appear to have an advantage over bottlefed children in developmental scales, IQ tests, and assessments of other specific cognitive outcomes. Among the most provocative observations are the positive effects on IQ of feeding human milk to premature infants. Although the workshop participants acknowledged controversial aspects of those observations, the need to replicate such studies was recognized widely. The consistency of the evidence argues strongly for evaluations with more robust designs and evaluation tools. Such investigations should permit inferences regarding the nature, degree, and persistence of the potential effects of breastfeeding or human milk feeding on behavioural development and the assessment of the modulation of putative effects by social, economic, and other environmental factors.


Benefits to the mother

Maternal benefits also fall into three broad categories: reductions in fertility, health benefits of a non-behavioural nature, and positive behavioural outcomes. The Working Group examined the first and second categories in greater detail than the third.

Generally, lactation is expected to help women maintain a healthy body weight when sufficient quantities of adequate food are readily accessible and to enhance the physiological efficiency of nutrient utilization under nearly all conditions. The hormonal changes that accompany lactation are expected to influence maternal behaviour in ways that support breastfeeding and promote mothering behaviours. Investigators also have suggested that successful breastfeeding is important to maternal self-efficacy and possibly social empowerment. These expectations likely are most relevant when maternal nutritional and social needs are met.

The mother's responses to breastfeeding have been studied much less than those of the infant. A principal limitation is that lactation seldom has been studied in the context of the complete reproductive cycle, which includes the nulliparous period, pregnancy, lactation, and the nonlactating, non-pregnant state that precedes a subsequent pregnancy. The significance of this omission stems from the likelihood that biological strategies for maintaining maternal well-being through the life cycle rely on a healthy physiological preparation for reproduction and adequate pregnancy intervals for maternal repletion. Interactions among the contiguous and interdependent stages within reproductive cycles and the biological effects of the distinct socio-economic, demographic, and environmental conditions in industrializing, newly industrialized, and post-industrialized settings are expected to modulate maternal responses to lactation.

Health consequences

Insufficient data were available to the Working Group for an assessment of the effects of lactation on the prevention of maternal obesity and nutrient depletion. Although obesity is of most concern in fully industrialized and newly industrialized nations, it is, ironically, a growing problem among some developing countries with large numbers of undernourished women of reproductive age. Similarly, the paucity of data makes it difficult to assess the global impact of lactation on nutrient depletion of the mother and its potential consequences for maternal and infant health.

Issues related to longer-term health outcomes, that is, osteoporosis and breast cancer, were addressed more confidently. Concerns that lactating women may be at greater risk of osteoporosis because of loss of calcium in milk have not been supported by recent studies conducted largely in affluent countries. Current evidence supports a preventive effect of breastfeeding against pre-menopausal breast cancer, but no association has been found between breastfeeding and post-menopausal disease.


Data reviewed by the Working Group reaffirmed the suppression of fertility by breastfeeding. The duration of the mother's infertility is directly dependent on her infant's suckling activity. Breastfeeding is most effective in decreasing fertility (and thereby facilitating longer, more desirable interpregnancy intervals) when infants are breastfed on demand and are provided no other sources of food or water. There also are data suggesting that the use of pacitiers may lessen the effects of breastfeeding on fertility by decreasing the infant's suckling activity.

The mean anovulatory period for non-breastfeeding women appears to be approximately 50 days. In breastfeeding women, anovulation may persist well into the second year post-partum. Infertility appears to be maintained by a suckling-induced disruption of the normal pulsatile pattern of luteinizing hormone (LH) release (essential for ovulation) and facilitated by an increased hypothalamic sensitivity to the negative feedback effects of oestradiol. The mechanisms responsible for these maternal responses to lactation have not been identified but are sufficiently reliable to have led a group of investigators to conclude that when women fully or nearly fully breastfeed and remain amenorrhoeic, breastfeeding provides more than 98% protection from pregnancy in the first six months post-partum. The programmatic implementation of this conclusion is known as the lactational amenorrhea method (LAM) of natural family planning.

Any biological or social factor that either promotes or interferes with the infant's suckling activity (such as a delay in the introduction of complementary foods or the inappropriate or premature introduction of supplementary or complementary infant foods) will, respectively, prolong or shorten the duration of infertility. Discussions on the control of milk synthesis were particularly relevant to these considerations. It is clear that milk synthesis is under autocrine (local) control. The frequency and degree to which the breast is emptied are the principal determinants of the quantity of milk that is produced. Generally, interference with the suckling activity of infants will be reinforced by a subsequent decrease in milk production. Under such conditions, feedback mechanisms will lead to progressive decreases in suckling, which, in turn, will disable mechanisms that disrupt pulsatile release of LH and eventually result in an earlier return of ovulation.

Demographic effects of breastfeeding

The Working Group examined the demographic effects of the impact of breastfeeding on fertility and infant mortality. It reviewed the impact of breastfeeding on one of the two principal proximate determinants of fertility, the rate of births. The other proximate determinant, the reproductive span (the interval between a woman's first ovulation and the time she either dies or becomes infertile), was not considered, because breastfeeding is not thought to influence it.

The effects of breastfeeding on the dynamics of birth intervals may be examined by dividing the birth interval into three parts: the post-partum period (the time between delivery and the resumption of both ovulation and sexual intercourse), the time between the end of the post-partum period and the next birth, and the period of the pregnancy associated with a live birth. Endocrine responses that make lactation possible prolong post-partum anovulation and amenorrhoea through mechanisms that have been reviewed briefly in the preceding section and regulate other reproductive functions (such as luteal function) through mechanisms that are understood less comprehensively. A semi-quantitative assessment of the impact of these effects on fertility suggests that a woman's lifetime fertility may be reduced as much as 50% by prolonged breastfeeding.

The demographic impact, however, also will be influenced by the effect of infant-feeding practices on child survival. Unlike the semi-quantitative assessments of the effects of the proximate determinants of fertility on population growth, the impact of the proximate determinants of child mortality on population growth has been more difficult to estimate. Six types of factors have been identified among the principal determinants of child mortality: maternal characteristics, environmental contamination, nutrient deficiency, injury, personal illness control, and the gestational age and development of the newborn. The first three are influenced greatly by breastfeeding.

The relations among breastfeeding, fertility, and child mortality are confounded by the socio-economic changes that often accompany changes in breastfeeding patterns. The socio-economic conditions that traditionally have led to decreases in the incidence and duration of breastfeeding tend, over the long term, to have a beneficial effect on the six factors listed above and to diversify and increase the use of contraceptive strategies for birth control. Nonetheless, if socio-economic changes are ignored and the positive impact of breastfeeding on fertility and child mortality on population growth is assessed, it appears that long-term breastfeeding (i.e., breastfeeding into the second year of the child's life) is likely to have only a limited effect on population growth. This, however, does not diminish the health benefits to both mother and infant anticipated from increased birth spacing and the nutritional and immunologic benefits discussed previously.


Current worldwide breastfeeding trends

The Working Group also reviewed data from demographic and health surveys conducted from 1990 to 1993. It is alarming that under-five mortality remains excessive by any measure in much of the world. For example, in 13 African countries for which data are available, mortality among children between one and four years of age ranged from 318 per 1,000 live births in Niger to 83 per 1,000 in Namibia. As in all regions, infant mortality in those 13 countries generally accounts for an increasing proportion of underfive mortality as the under-five mortality rate drops.

It is very likely that improved breastfeeding practices will have a significant impact on child mortality in nearly all economically developing countries. The term "breastfeeding practices" deserves emphasis, because the percentage of children born in the last five years who were ever breastfed ranged from 95% to 97% in the same 13 African countries for which mortality data are available. The percentages of children ever breastfed were similarly high (greater than 90%) in the Asian, South Pacific, and Latin American countries for which data are available.

In most developing countries that were surveyed, substantially more than 50% of all infants were breastfed up to 12 to 15 months of age, and more than 25% were breastfed up to 20 to 23 months. The median duration of breastfeeding among children born in the last three years ranged from 17 to 28 months in the African countries that were surveyed. No economically developing country in the regions surveyed had a mean duration of breastfeeding below six months, and in most countries the mean durations were substantially above that level. Yet, consistently across all countries surveyed, the mean duration of breastfeeding was from 5% to nearly 100% greater in rural than in urban areas. In most countries, a minority of infants were fed only human milk through four months of age, although rates varied widely among those countries surveyed. For example, 90% of Rwandan infants were reported to receive only human milk through four months of age. The rates in Tanzania, Kenya, Madagascar, and Namibia ranged from 17% to 47%, and the rates in Burkina Faso, Ghana, Malawi, Niger, Senegal, Nigeria, Zambia, and Cameroon ranged from 1% to 13%. Rates were similarly divergent in other regions of the world. The percentages of infants whose diets were restricted to only human milk and water were similarly divergent among countries but were substantially higher than the percentages of those receiving only human milk.


Sociocultural factors affecting breastfeeding

Breastfeeding is a learned, not an instinctive, behaviour. Desirable breastfeeding practices must be actively promoted and supported. Successful breastfeeding, therefore, is dependent upon social and cultural factors. Major shifts in breastfeeding practices in fully industrialized countries over the last 30 to 40 years and rural-urban differences in most economically developing countries provide the best evidence of the great influence of sociocultural factors on breastfeeding. The best predictors of breastfeeding practices in fully industrialized countries are sociocultural rather than biological. This also is increasingly true in the industrializing countries, especially in those that are urbanizing quickly. However, recognizing the importance of sociocultural factors in determining infantfeeding practices does not lessen the difficulty of understanding how specific sociocultural factors operate or may be measured adequately to explain variations within and between different infant-feeding patterns.

The sociocultural factors that have been examined most often are those that can be integrated easily into biomedical and epidemiologic models, such as religion, martial status, education, and kinship patern. These often are included in assessments of knowledge, attitudes, and beliefs. Yet because infant feeding, and breastfeeding in particular, represents a wide range of highly emotional issues, it is often difficult to obtain reliable and valid data from informants in most studies.

Other factors are less commonly studied, because they are more difficult to assess. For example, factors reflective of values, attachment, nurturance, and sexuality require interpretation from social science paradigms and are not as amenable to reductionist models. Nonetheless, all of these factors probably contribute significantly to the links among what people say they know, what they know, and what they practice.

As long-term, detailed ethnographic analyses have become increasingly available, a conceptual model has emerged that describes culture as an interaction between style and structure. Style refers to the manner of expression characteristic of an individual, a time, and a place. The application of this model is expected to increase understanding of the influence of sociocultural factors on breastfeeding. Infant-feeding styles communicate fundamental values, attitudes, and beliefs reflected in the interaction between mother and infant during feeding, in how breastfeeding is accomplished, and so forth. These styles of feeding are part of dynamic trends and fashions.

Styles in turn are in a dynamic interaction with defined organizational and institutional structures, such as those related to health care, the economy, and governments, each with its own potential influence on infant-feeding choices. An improved understanding of relevant styles and structures should enhance our ability to predict how infant-feeding choices will be affected by changes in sociocultural factors.

Despite these limitations, a comparison of the effects of biological and sociocultural factors on measures of breastfeeding success (for example, prevalence and duration) strongly suggests that breastfeeding is biologically robust but highly susceptible to positive and negative sociocultural influences. The principal basis for this conclusion is that breastfeeding is sustainable under the wide range of biological conditions characteristic of affluent women in economically developed countries and of poor women in harsh environments in less economically developed areas. This is not true when breastfeeding is considered under an analogously wide range of sociocultural conditions relevant to breastfeeding. Although it would be a mistake not to recognize the cost that this characteristic presents to poor women (that is, to their biological wellbeing), it is equally fallacious to conclude that adequate breastfeeding can be accomplished only when all biological needs are optimally met.


Resources needed to protect, support, and promote breastfeeding

The information reviewed by the Working Group did not allow a prioritization of resources needed to protect, support, and promote breastfeeding. It did allow the group, however, to identify resources that would enhance the likelihood of successful lactation in nearly all settings. The paucity of quantitative information available to assess the relative importance of resources needed in specific settings represents a major research gap. The resources identified by the group fell into three broad categories: time, space, and sociocultural/economic support.

The physiological and sociocultural information reviewed by the group documented clearly that breastfeeding requires time of the mother. The two principal sources and sinks of time are the family and, when the mother also is employed outside the home, her employer. Because milk production is sustained by physiological processes dependent upon the regular removal of milk, time constraints that result in decreased or inefficient suckling will have a negative impact on milk production and eventually on the sustainability of adequate milk production. Time constraints imposed by employers have marked negative impacts on breastfeeding success because of adverse effects on suckling. Employment policies that recognize the importance of maternal leaves, temporary parttime employment options that do not adversely affect longer-term full-time employment opportunities, and opportunities for breastfeeding in the workplace represent complementary strategies to help establish and sustain adequate lactation.

Space is required to breastfeed. Differing perceptions of physical modesty, hygiene, and other concepts dependent upon cultural norms and relevant to infant feeding and maternal well-being will make diverse demands on the characteristics of spaces best suited for the protection, support, and promotion of breastfeeding. These demands apply to family residences, places of employment, and various sites where communities congregate, such as places of worship, businesses, and entertainment.

Sociocultural and economic support fall into two subcategories, tangible and intangible support. Examples of the types of tangible support needed to obtain full benefits of breastfeeding are safe and adequate food for the mother and complementary infant foods for the period of mixed feeding when foods other than human milk are introduced to the infant's diet; fair labour compensation that recognizes the needs of families; and adequate housing and related services that protect, support, and promote the hygienic well-being of the family.

Examples of intangible support tended to centre around five social sectors: government, business, community, health professions, and educational and research institutions. Those which centre around government represent a wide range of issues. They extend from laws and policies that govern parental leaves to those that lead to differing urbanization trends. Parental leave policies are of obvious relevance; urbanization trends influence family support structures and employment patterns, which affect the protection, support, and promotion of breastfeeding.

Although the Working Group recognized the significant influence that the commercial sector plays in determining parental and family leave policies of specific countries, the negative impact of both overt and subtle inappropriate marketing practices by producers of infant foods received more focus. Strategies that have a negative impact on breastfeeding appear designed to decrease suckling at the breast, thereby causing decreased milk production, with increased dependence on human milk substitutes, and undermining maternal confidence in the ability to breastfeed and the general social support of breastfeeding. These strategies are implemented by such diverse activities as direct advertisement to the public and the now discredited distribution of human milk substitutes at little or no cost in health-care settings or directly to family residences. Other issues relevant to the commercial sector's employment policies and the impact of these policies on the time mothers have to breastfeed have been discussed previously.

The issue common to communities-at-large, health professions, and educational and research institutions is recognition of breastfeeding as the expected mode of feeding for all infants and, its corollary, the use of human milk substitutes only when specifically indicated. Although all agencies and institutions with interests in infant health recommend exclusive breastfeeding for at least the first four to six months, these recommendations are not commonly reflected in the practices of communities, health professionals, and educational and research institutions. Examples of the consequences of failing to make practices conform with recommendations are inappropriate management of lactation by health professionals who have received inadequate training, poor knowledge and attitudes of many young families relative to breastfeeding because of inattention to lactation in primary and secondary education, and a poor knowledge base for the improvement of lactation practices because of inadequate research support.


Conclusions and recommendations

The data reviewed by the Working Group provide a strong scientific base for the present recommendations. The present benefits of breastfeeding in all countries and the benefits that are projected when international recommendations are implemented more broadly are of great significance to individuals and organizations responsible for the implementation of scientific knowledge that is highly pertinent to infant and maternal health.

The Working Group urges the active protection, support, and promotion of breastfeeding by governments, communities, the commercial sector, educational and research institutions, voluntary organizations responsible for the promotion of maternal and infant health, and, in particular, health professionals and facilities. Especially relevant to this recommendation is the resilience of lactation in the face of harsh biological conditions and the fragility of breastfeeding in the face of inadequate sociocultural and economic support. These characteristics impose a special responsibility on all societies to safeguard the well-being of women by ensuring their access to a safe and adequate food supply throughout their life cycle and to provide adequate time, space, and sociocultural and economic support to women and their families to maximize the health of all children from infancy and the health of women throughout the reproductive cycle.

Contents - Previous - Next